but 'wanting' or 'hoping' and the dice roll isn't the important part of the action resolution of what makes them different from each other, attacking the orc is a closed-potential outcome, there isn't any possibilities beyond 'hit' or 'miss' when you attempt to resolve the action.
Is it? The possibility of things like critical hits, fumbles, narration resulting in a changed situation of battle, altered priorities, etc. mean that I would not call it a totally closed form. Yes, it is
closer to close-ended, but it is not totally so. Otherwise, there would--could--be one and only one narration for what rolling to hit meant on success or failure.
deciphering the runes in this situation is not the same case, there is ambiguity to their meaning, if the runes example was the same as the attack example it would be a question of 'can i translate the runes? yes or no'
Except that IS closed-form. There can only be one answer to that; yes, you can, or no, you can't. You have actually made that
less open than even an attack roll.
and the GM would know their meaning and tell you on a success,
You have presumed GM pre-authorship by doing this. In other words, you have
presumed the very thing you're trying to demonstrate something about. That's a pretty serious weakness for this argument.
and if the attack example was the same as the runes example it would a question of 'does my attack attempt work? yes or no' but then on a success you could list any type of offensive action against the target,
Surely we do in fact allow such things though? Like if someone wants to trip their target, that's going to be an attack roll, possibly with some kind of modifier to affect the success. Presumably, to respect the specialization of Battle Masters who actually do
have a specific tripping maneuver, they would forgo their damage in order to get this benefit. (That's how I've run a number of similar such things: trained folks get their goodies AND their damage, the untrained get their goodies OR damage.) And that exact thing is precisely how Battle Master maneuvers work; the vast majority of them are something you declare
after you have rolled to-hit, not before.
you could damage them, you could inflict status, you could even use magic against them because your 'nonspecific attack action' was successful.
Is that not how a Battle Master with some kind of magic training works? Consider a Ellaria, an Elf Fighter (Battle Master) 3/Wizard (Bladesinger) 6. With a single attack roll, she can elect to
either perform any maneuver she knows (so long as she still has Superiority Dice left)
or any of her cantrips that require an attack roll, since a 6th level Bladesinger can replace one of their two attacks from the Attack action with a cantrip of their choosing.
Your claim is that an attack roll is necessarily one, and only one, specific action, every single time. That isn't true in general, because players often attempt a variety of things for which "an attack roll" is the best-fit mechanic. It also isn't true in several common but specific cases, like Battle Masters, who can (indeed, usually
must) declare,
after hitting with an attack, that they were using a specific maneuver. And it isn't true of Bladesingers (I just made the example more pointed by using a Battle Master multiclass), who can blend attack cantrips with their regular attack rolls just fine.
So even with this allegedly basic mechanic, we can see that it actually has quite an open-ended spectrum! That would seem to lend a lot of credence--not necessarily a smoking gun, but perhaps the smell of spent powder--to gban007's argument that there
is an important symmetry between the "can I read the runes, hoping they will be X?" and "can I attack the orc, hoping I will achieve X?"